How Mistral Data is transforming public transport

I recently interviewed Duncan Waugh, managing director at Mistral Data, a UK-based company that is working to improve public transport using data and services built upon Amazon Web Services (AWS). In this blog, you’ll learn how they are helping a traditional industry innovate, improving operations and the customer experience.

Duncan, can you tell me more about Mistral Data?

We are a relatively new company formed out of FirstGroup, a market leader in the public transport industry. Our aim is to provide data-driven software to both the UK and international transport industries using AWS cloud technology.

Our business is all about the use of data to provide quicker and clearer insight, allowing better decision-making to be made. We have focused on those areas where there is the greatest opportunity to make a difference in the customer experience, which is running trains better and keeping our customers more informed before and during their journeys.

We currently offer 14 different products, and we are looking to increase the capability and features of these existing products, develop new products and services with our development partners in the UK, Europe, and India, and extend our customer base.

How do those products improve the customer experience when using public transport?

Public transport is a traditional industry that has a mix of old, infrequent data going back 60 years and more modern data sources that offer much greater granularity but aren’t yet integrated with these older data sources. Traditionally, data is siloed within individual functions for a specific purpose. Our approach is to use AWS services to integrate data from multiple sources and create a single view of operational and customer data to share with colleagues and customers through different services to help them make better decisions in using the railway.

A few examples of the services we’ve created using AWS are:

Berth Maps. The rail industry uses a number of schematic map applications that show train locations based on data generated from Network Rail. We complement this data with GPS-generated data and map these to signals and berths, creating a “virtual berth,” which provides a more granular view of each train’s location and a replay capability so that train movement can be compared with signal aspect for remediation planning.

This product helps monitor rail performance. Because it is built using AWS native services, it is scalable, resilient, and can be rolled out quickly to any part of the rail network when the map and signal location are created. The cost is 40–60 percent less than other offerings, with all services accessed through the web.

Notus. We have recently launched a proof of concept with one of our customers, where we ingest a real-time data stream generated from the black box recorders on each train and use changes in train behavior, such as wheel slip or braking, to identify where trains are experiencing low adhesion, which makes it possible for us to inform drivers of problem areas, and also where remediation activity will be taking place, to improve the performance of the train service, especially during autumn when falling leaves can cause slippery tracks and delays.

This product uses AWS native services, including

• Amazon Kinesis Streams, a service to easily stream data at any scale,
• Amazon DynamoDB, a fast and flexible NoSQL database service for any scale,
• AWS Lambda, which makes it possible for organizations to run code without thinking about servers or clusters,
• Amazon API Gateway, a service to create, maintain, and secure APIs at any scale
• Amazon Simple Notification Service (Amazon SNS), a fully managed Pub/Sub service for A2A and A2P messaging, and
• Amazon Simple Queue Service (Amazon SQS), a fully managed message queuing for microservices, distributed systems, and serverless applications

to apply business rules to a real-time data stream and then present that in a user-friendly format to improve the performance of the railway.

Mistral API. We present all our data from our single view of train operations to various downstream channels (e.g., colleague and customer digital channels such as the app and website that show where the train is, its coaches and orientation, the facilities in each coach, and how busy each vehicle is). We recently tuned the database to improve performance sixfold.

How are you using technology to empower your customers and employees?

We are changing the way data are used within the rail industry to help make decisions faster and to improve the performance of the railways and the customer experience by providing detailed, accurate, and timely information about their journey.

We will shortly be integrating the single view of train operations I mentioned, with our single view of customers to provide personalized alerts and messages to customers to help answer questions like “Where do I stand on the platform to find my seat?” or “Has my train had its timetable changed?” We developed this such that the relevant data is being sent to a large email and push notification marketing engine, but we have also deployed email and push notifiations using AWS Pinpoint—which offers organizations one customizable tool to deliver customer communications across channels, segments, and campaigns at scale—for one of our customers recently. This has reduced our annual costs by 65 percent for the software alone, with support being much easier and cheaper to provide, whereas the reporting and monitoring have a much greater capability. We are looking to extend this service to other customers.

We’re also rolling this message orchestration concept out in an app for our frontline colleagues. They will be able to receive a personalized alert relating to the train service they’re working for or from the control center. Staff can also submit information in near real time relating to train defects or delay reasons.

All this is managed through AWS services—a mixture of Amazon Kinesis Data Streams, AWS Lambda, and Amazon Relational Database Service (Amazon RDS), which is a collection of managed services that make it simple to set up, operate, and scale databases in the cloud (Postgres and Aurora databases) as well as Amazon Pinpoint for messaging and Amazon SageMaker—which helps customers to build, train, and deploy machine learning (ML) models for any use case with fully managed infrastructure, tools, and workflows—for ML features. All data are collected in near real time and then pushed to a database on Amazon Redshift—a service used to analyze structured and semi-structured data across data warehouses, operational databases, and data lakes—to facilitate trends and analytics using a range of visualization tools from AWS QuickSight, a popular cloud-native, serverless BI service, to Power BI and Sigma.

You mentioned that Mistal Data is all about improving operations and the customer experience using data. Can you tell us about other ways you are using data?

Yes, we have built revenue management systems to optimize pricing for tickets bought in advance. We are trialing the use of ML to better dynamically segment customers and to better market to them based on their previous travel behavior. Using control center data, we can provide views of what is happening across the rail network at any moment, including areas of low adhesion challenges. And we can support all parts of a train operator’s business with data that are ingested from 40 source systems to facilitate business intelligence reporting and analytics.

How has your collaboration with the AWS team helped you to deliver this level of innovation to your employees and customers?

Around 5 years ago, First Group’s Rail Division decided to pursue a data strategy using AWS. At the time, the focus was as much on resilience and cost avoidance for new servers and hardware. However, it quickly became clear to us that the biggest advantages were around the speed with which new ideas could be tested and implemented. The use of serverless technology has massively reduced our IT management overhead and helped us instead to focus on developing these new data-driven solutions while ensuring we have scalability, resilience, performance, and security of all data that is held within our platforms. The additional services we have started to use, such as Amazon Pinpoint and Amazon SageMaker, are demonstrating value at much lower prices than other typical applications that provide these services.

Without AWS, we would not have launched Mistral Data, as we would not have been able to develop many of the products we now sell and support, due to the cost and risk involved with innovation. The support from AWS has been incredibly proactive in encouraging us to use different services and providing technical expertise to help us do that.

AWS has made every effort to help us be successful, continually emphasizing the focus on customer needs and how to meet them.

Visit www.mistral-data.com to learn more.